MMN distributed sources. Evidence from human intracranial recordings

BLENKMANN ALEJANDRO; PHILLIPS HOLLY; CHENNU SRIVAS; ROWE JAMES; MURAVCHIK CARLOS; KOCHEN SILVIA; BEKINSCHTEIN TRISTAN

Washington DC

Encuentro; SFN 2015 annual meeting; 2015

Society for Neuroscience

*Trabajo invitado en el segmento - IBRO Travel Award Winners-Idem a trabajo presentado en 7th Mismatch Negativity Conference, Leipzig, Alemania.OBJECTIVES: In order to investigate MMN sources we recorded electrocorticography (ECoG) and/or Stereo-electroencephalography (SEEG) from human epilepsy patients during three different MMN paradigms.METHODS:Fourteen epilepsy patients implanted with grids and depth electrodes were studied. View table for implantation details.Image acquisition and processingPost implantation T1 MRI (grid patients) or Pre implantation T1 MRI (depth electrodes patients) Post implantation CTImage corregistration and normalization to MNI152 standard. Electrodes localization using iElectrodes toolbox (Blenkmann et al 2014, Blenkmann et al, 2015).We used three paradigms to evoke MMN potential:MMNa: An auditory MMN Optimum-1 paradigm alternating standard tones (S) with deviant tones (D#) differing by frequency, intensity, location, duration or a silent gap (Näätänen et al. 2004). 600 tones for each block. Each tone lasts 100ms, SOA = 500ms. 3 blocks, lasting ~15 minutesMMNo: Four series of 5 complex tones were used. Two using the same 5 sounds (AAAAA or BBBBB) and two with the final sound swapped (AAAAB or BBBBA) (Bekinschtein et al, 2009).Interaural MMN: A modification of the previous paradigm (Chennu et al 2013), where sequences were either monaural or interaural (8 blocks, lasting ~ 90 min). Tree runs with different tasks:MMNton: Subjects were asked to count deviant tonesMMNseq: Subjects were asked to count deviant sequencesMMNvis: Subjects were asked to count color letters in a computer screen (uncorrelated to sound onset)EEG acquisition and preprocessing2000Hz or 2000Hz downsampled to 200Hz.Filtering 1-40Hz (zero phase)Average neighbor re-reference to improve signal from local sources compared to distant ones.Channels with epileptic activity or noise were rejectedEpoching -100 to 400 ms from tone onsetTrials with more than 3 std deviation were rejected and visually checked. MMN potentials were studied in the 75-225 ms time window. Permutation tests were conducted to assess statistical significance of ERP differences for at least 25 ms (p